Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 14 Dec 2021 00:15:01 -0800
From:      Mark Millard via freebsd-ports <freebsd-ports@freebsd.org>
To:        "ler@freebsd.org" <ler@FreeBSD.org>
Cc:        freebsd-ports@freebsd.org
Subject:   Re: lang/go: sigbus on 13-Stable Jail?
Message-ID:  <A8B9CD27-4AEB-4D8A-9458-79A6CEDE286C@yahoo.com>
In-Reply-To: <3D942FF9-F517-42AB-8C25-027F696913AF@yahoo.com>
References:  <3D942FF9-F517-42AB-8C25-027F696913AF@yahoo.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On 2021-Dec-13, at 23:16, Mark Millard <marklmi@yahoo.com> wrote:

> From: Larry Rosenman <ler_at_FreeBSD.org> wrote on
> Date: Mon, 13 Dec 2021 18:48:52 -0600 :
>=20
>> =
https://home.lerctr.org:8888/data/p130-S-amd64-host-ports/2021-12-13_18h14=
m58s/logs/errors/go-1.17.5,1.log
>>=20
>>=20
>> # cmd/compile/internal/ssa
>> fatal error: unexpected signal during runtime execution
>> [signal SIGBUS: bus error code=3D0xc addr=3D0x8198d4000 pc=3D0x429121]
>>=20
>> runtime stack:
>> runtime.throw(0xc2702b, 0x2a)
>> 	/usr/local/go/src/runtime/panic.go:1112 +0x72
>> runtime.sigpanic()
>> 	/usr/local/go/src/runtime/signal_unix.go:671 +0x46a
>> runtime.(*pageAlloc).alloc(0x1206c48, 0x8, 0x0, 0x0)
>> 	/usr/local/go/src/runtime/mpagealloc.go:767 +0x81
>> runtime.(*mheap).allocSpan(0x1206c40, 0x8, 0x420000, 0x121c5a8, 0x48)
>> 	/usr/local/go/src/runtime/mheap.go:1122 +0x62f
>> runtime.(*mheap).alloc.func1()
>> 	/usr/local/go/src/runtime/mheap.go:871 +0x64
>> runtime.(*mheap).alloc(0x1206c40, 0x8, 0x460100, 0x1206c40)
>> 	/usr/local/go/src/runtime/mheap.go:865 +0x81
>> runtime.largeAlloc(0xf8a0, 0x1210001, 0x882348ae0)
>> 	/usr/local/go/src/runtime/malloc.go:1152 +0x92
>> runtime.mallocgc.func1()
>> 	/usr/local/go/src/runtime/malloc.go:1047 +0x46
>> runtime.systemstack(0x0)
>> 	/usr/local/go/src/runtime/asm_amd64.s:370 +0x66
>> runtime.mstart()
>> 	/usr/local/go/src/runtime/proc.go:1041
>>=20
>>=20
>> Ideas?
>>=20
>> Today's 13-Stable, and today's ports tree
>> (Poudriere hosted on 14-Current)
>=20
> I tried building lang/go for stable/13 (from main) via
> poudreire-devel in/for a somewhat older amd64 context:
> (from the build's log file)
>=20
> Ports top last git commit: 9bed79e86972
> Ports top unclean checkout: yes
> Port dir last git commit: 66b73ca2e1fa
> Port dir unclean checkout: no
> Poudriere version: poudriere-git-3.3.99.20211130
> Host OSVERSION: 1400043
> Jail OSVERSION: 1300522
>=20
> For reference:
>=20
> # cd /usr/ports/
> # ~/fbsd-based-on-what-commit.sh=20
> branch: main
> merge-base: 9bed79e869721b4ca8a15c527db8d40969867c2c
> merge-base: CommitDate: 2021-12-06 03:45:47 +0000
> 9bed79e86972 (HEAD -> main, freebsd/main, freebsd/HEAD) www/drupal9: =
update to 9.2.10
> n567671 (--first-parent --count for merge-base)
>=20
> # cd /usr/13S-src/
> # ~/fbsd-based-on-what-commit.sh
> branch: stable/13
> merge-base: c89c8b894cffd033d6409f7bf0bdb7543bbf78f9
> merge-base: CommitDate: 2021-12-07 18:53:33 +0000
> c89c8b894cff (HEAD -> stable/13, freebsd/stable/13) net80211(4): Fix a =
typo in a sysctl description
> n248443 (--first-parent --count for merge-base)
>=20
> # cd /usr/main-src/
> # ~/fbsd-based-on-what-commit.sh
> branch: main
> merge-base: 22c4ab6cb015dc99eb82504e5fd957662cded3c3
> merge-base: CommitDate: 2021-12-07 19:29:26 +0000
> 22c4ab6cb015 (HEAD -> main, freebsd/main, freebsd/HEAD) sys/_bitset.h: =
Fix fall-out from commit 5e04571cf3c
> n251456 (--first-parent --count for merge-base)
>=20
> lang/go is older as a consequence: 1.17.4,1
>=20
> No CCACHE use in my context, unlike yours.
>=20
> The log shows a smaller stack size limit: stack size              =
(kbytes, -s)  524288
> The log shows a larger open files limit : open files                   =
   (-n)  4096
>=20
> Result:
>=20
> [00:04:23] [01] [00:03:01] Finished lang/go | go-1.17.4,1: Success
>=20
>=20
>=20
> (This happened to separately be a useful test of a newly set
> up part of my environment. But my context being only about a
> week back, I figured that the status for such a short time
> back might prove useful.)

I should have also reported about the jail used:

# poudriere jail -j13S-amd64-bulk_a -i
Jail name:         13S-amd64-bulk_a
Jail version:      13.0-STABLE
Jail arch:         amd64
Jail method:       null
Jail mount:        /usr/obj/DESTDIRs/13S-amd64-poud-bulk_a
Jail fs:          =20
Jail updated:      2021-12-13 22:11:32
Jail pkgbase:      disabled

/usr/obj/DESTDIRs/13S-amd64-poud-bulk_a was actually initialized
from inside a chroot to a stable/13 area --and was filled-in via
using

. . . installworld distrib-dirs distribution DB_FROM_SRC=3D1 . . .

for later use by poudriere-devel. (This style of use allows me to
install and use the same stable/13 build that I install for my
normal (direct) stable/13 use.)

So, while main used the stable/13 jail, main did not build
the world in use by that jail.

(I use ZFS in order to use bectl :

# bectl list
BE          Active Mountpoint Space Created
13S-amd64   -      -          4.79G 2021-08-20 16:57
13_0R-amd64 -      -          4.83G 2021-08-20 16:56
main-amd64  NR     /          6.03G 2021-08-20 16:57

. It is possible to chroot into one of the 13*'s from
having booted main, based on being able to mount the
one(s) of interest first. I also have other
mount_nullfs's in use for such a chroot to be able to
use various areas that are shared across the contexts.)

=3D=3D=3D
Mark Millard
marklmi at yahoo.com
( dsl-only.net went
away in early 2018-Mar)




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?A8B9CD27-4AEB-4D8A-9458-79A6CEDE286C>